Abstract
Background and Aims
Underwater endoscopic mucosal resection (UEMR) has been widely adopted for the treatment of superficial nonampullary duodenal epithelial tumors (SNADETs). The aim of this study was to identify predictors of local recurrence and piecemeal resection in UEMR for SNADETs.
Methods
This is a retrospective single-center study. Patients who underwent UEMR for SNADETs between November 2017 and October 2024 were included. We documented clinicopathologic characteristics of patients and lesions, noted procedure-related outcomes, and evaluated local recurrence. Kaplan-Meier analysis was used to assess recurrence-free survival, and logistic regression analysis was performed to identify the predictors of piecemeal resection.
Results
Among the 367 patients, the rate of piecemeal resection was 11.2%, and the rate of en bloc resection with histologically tumor-free horizontal and vertical margins (R0) was 65.1%. No intraoperative perforations occurred; intraoperative bleeding was observed in 1.9%. Kaplan-Meier analysis of recurrence-free survival showed significantly poorer outcomes in the piecemeal resection group than the en bloc resection group (log-rank test, P < .001; Wilcoxon test, P < .001). According to multivariate logistic regression analysis, each 1-mm increase in lesion size (odds ratio [OR], 1.20; 95% confidence interval [CI], 1.11-1.31; P < .001) and a flat macroscopic type (OR, 12.49; 95% CI, 1.45-107.98; P = .02) were significant independent predictors of piecemeal resection.
Conclusions
Piecemeal resection was a risk factor for local recurrence in UEMR for SNADETs. In addition, each 1-mm increase in lesion size and a flat macroscopic type were significant independent predictors of piecemeal resection.
Introduction
Superficial nonampullary duodenal epithelial tumors (SNADETs), traditionally considered rare,1,2 have been detected more frequently as a result of advances in endoscopic imaging technology and increased awareness in clinical practice.3 Because SNADETs have malignant potential,4 endoscopic resection must be performed early and effectively. However, the optimal endoscopic resection strategy remains controversial because each technique has different inherent advantages and limitations.5
Conventional endoscopic mucosal resection (CEMR) has been widely used for the treatment of SNADETs because of its technical simplicity and favorable safety profile.6, 7, 8, 9, 10 However, duodenal CEMR may sometimes be inadequate for lesions with submucosal fibrosis induced by preoperative biopsy.11,12 For large lesions, endoscopic submucosal dissection (ESD) enables en bloc resection with high reliability, but it is technically demanding and associated with a higher incidence of adverse events, such as perforation and bleeding.13,14
Underwater endoscopic mucosal resection (UEMR) for SNADETs was first reported, to our knowledge, by Binmoeller et al15 in 2013 and has since been widely adopted as a therapeutic approach for SNADETs.16,17 Because UEMR eliminates the need for submucosal injection, the procedure is simpler and shorter than CEMR.18,19 Furthermore, UEMR is less hampered by submucosal fibrosis, which represents a significant advantage.12,20
Despite the growing adoption of UEMR, the predictors of local recurrence and piecemeal resection remain poorly defined. Piecemeal resection challenges the accuracy of pathologic evaluation21 and is associated with an increased risk of local recurrence.6,8,13,22, 23, 24, 25 Identifying the predictors of piecemeal resection is essential for optimizing patient selection, refining procedural techniques, and improving clinical outcomes. We aimed to identify the independent predictors of local recurrence and piecemeal resection in UEMR for SNADETs through a comprehensive analysis.
Methods
Study design and patient eligibility
In this retrospective observational study, participants were consecutive patients with SNADETs who underwent UEMR at our institution between November 2017 and October 2024. Patients with familial adenomatous polyposis, those who had previously undergone upper gastrointestinal surgery, and those who were undergoing treatment for locally recurrent lesions were excluded. Cases involving partial injection UEMR (PI-UEMR) were excluded to focus exclusively on standard UEMR procedures.
Indications for UEMR
We performed UEMR primarily for SNADETs measuring <20 mm in diameter. It was also performed for lesions measuring ≥20 mm that were pedunculated and considered amenable to en bloc resection with a snare. For lesions measuring ≤12 mm that were suspected of being low-grade adenomas with high certainty, cold snare polypectomy (CSP) was the preferred technique, in accordance with previous studies suggesting that 12 mm is an appropriate cutoff for CSP based on both carcinoma risk stratification and clinical outcomes.4,26 For lesions with significant fibrosis as a result of prior endoscopic treatment or flat lesions in the duodenal bulb that were nearly 20 mm in size, ESD was performed instead of UEMR if en bloc resection was deemed unfeasible.
UEMR procedure
All patients were hospitalized and underwent UEMR under conscious sedation induced by a combination of a benzodiazepine and pethidine hydrochloride. Periprocedural management of antithrombotic agents was conducted according to the Japanese guidelines established by the Japan Gastroenterological Endoscopy Society (JGES).27,28 Antithrombotic therapy was adjusted based on the type of agent and the patient's thromboembolic and bleeding risks. For example, direct oral anticoagulants were withheld on the morning of the procedure and resumed the next day if no delayed bleeding occurred. Antiplatelet therapy was continued, withheld, or substituted with aspirin or cilostazol depending on risk.
The procedures were performed by 28 endoscopists, of whom 21 were board-certified fellows of the JGES and 7 were noncertified. They used primarily a therapeutic endoscope with a water jet function: either the GIF-H290T or the PCF-H290TI (Olympus Medical Systems, Tokyo, Japan). The electrosurgical unit primarily used was VIO 3 (Erbe Elektromedizin, Tübingen, Germany). The operator’s choice of endoscopic snare was based on the lesion size: either SnareMaster snares (10, 15, or 20 mm; Olympus Medical Systems) or Captivator II snares (13 or 16 mm; Boston Scientific Japan, Tokyo, Japan). Lesion size was measured with open biopsy forceps (4 mm in diameter) or endoscopic measure (M2-3U; Olympus Medical Systems).
The UEMR procedure is illustrated in Figure 1. First, the air in the duodenal lumen was completely suctioned, and the endoscope's water jet function was used to fill the lumen with warm normal saline solution (Fig. 1A and B). Next, the operator used an endoscopic snare to capture the lesion, thereby ensuring an adequate safety margin (Fig. 1C). An electrosurgical current (endoCUT Q mode, effect 1.0) was used to resect the lesion. After resection, the site was examined for any residual tissue (Fig. 1D); if it was detected, the snare or biopsy forceps were used to perform additional resection. The mucosal defect was closed with endoscopic clips, either EZ Clip (Olympus Medical Systems) or SureClip (Micro-Tech, Nanjing, China). In our institution, en bloc resection was attempted in all cases. Lesions deemed unsuitable for en bloc resection with a snare based on preoperative assessment were treated with ESD instead of UEMR.
Figure 1.
Illustration of the underwater endoscopic mucosal resection procedure. A, Duodenal polyp (arrow) visualized in an air-filled duodenal lumen. B, The air within the duodenal lumen is completely suctioned, after which the lumen is filled with warm normal saline solution. C, An endoscopic snare is used to capture the lesion with an adequate margin. D, The lesion is resected with the use of an electrosurgical current, and the resection site is visualized endoscopically.
Measured outcomes
We documented the clinicopathologic characteristics of the patients (eg, age and sex) and of the lesions (including size, location, circumference, macroscopic type, and whether preoperative biopsy had been performed). In addition, we documented the treatment period (early [November 2017 to December 2020] and late [January 2021 to October 2024]) and the certification status of the endoscopists who performed the procedures. Procedure-related outcomes included rates of en bloc resection and en bloc resection with histologically tumor-free horizontal and vertical margins (R0), procedural time, histologic outcomes (including histologic type, grade, and resection margin status), adverse events such as intraoperative perforation and bleeding, and the presence or absence of local recurrence during follow-up. R0 was defined as en bloc resection with histologically tumor-free horizontal and vertical margins. R1 was defined as resection with positive horizontal and/or vertical margins, and RX was defined as resection with either the horizontal or vertical margin could not be assessed. R1 and RX were collectively defined as non-R0. In addition, all piecemeal resections were classified as non-R0. Procedural time was defined as the interval from the insertion of the snare through the working channel to the confirmation of complete lesion resection and absence of a residual tumor at the ulcer base; time required for clip closure was not included. Intraoperative perforation was defined as any perforation identified during UEMR. Intraoperative bleeding referred to bleeding that necessitated hemostasis with hemostatic forceps. Minor bleeding controlled with the snare tip or with clips used for mucosal closure was not considered intraoperative bleeding. Delayed bleeding was defined as any postprocedural bleeding that required endoscopic hemostasis or blood transfusion. Local recurrence was assessed during follow-up endoscopic surveillance, which was typically conducted at 6- to 12-month intervals after resection using magnifying endoscopy. Lesions with endoscopic features suggestive of recurrence were biopsied, and recurrence was defined as histologically confirmed neoplasia at the site of the original resection. Kaplan-Meier analysis was performed to evaluate recurrence-free survival according to resection method and histologic margin status. Differences between groups were assessed using the log-rank test and the Wilcoxon test.
Statistical analysis
To identify the factors associated with piecemeal resection and non-R0 (RX or R1), we performed both univariate and multivariate logistic regression analyses. Categorical variables are expressed as counts and percentages, and continuous variables are presented as mean (standard deviation). The following variables were considered potential risk factors: longitudinal lesion location (duodenal bulb vs other locations), lesion size, macroscopic type (flat type vs protruding type), history of preoperative biopsy, treatment period (early [November 2017 to December 2020] and late [January 2021 to October 2024]), and certification status of the endoscopists (board-certified fellows of the JGES and noncertified). In addition, we analyzed the receiver operating characteristic (ROC) curve to determine the optimal cutoff lesion size for predicting piecemeal resection.
For each outcome (piecemeal resection and non-R0), univariate logistic regression analysis was first performed, and variables with a P value of <.2 were considered candidates for inclusion in the multivariate models to avoid excluding potentially important variables. In addition, clinically relevant variables, such as lesion size, macroscopic classification, and certification status, were included in the multivariate models regardless of their univariate P values. In the model for non-R0, we also included preoperative biopsy as a covariate, despite its univariate P value exceeding .2, because it may influence histologic evaluation and resection margin assessment.
All statistical analyses were performed with JMP software (version 17.2.0; SAS Institute Inc, Cary, NC, USA). A P value of <.05 was considered statistically significant in the multivariate analysis, and statistical significance for all logistic regressions was assessed using the likelihood ratio test.
Results
Patient characteristics
The flowchart for inclusion in this study is shown in Figure 2. The baseline characteristics of the 367 participants are summarized in Table 1. Approximately 6% of the lesions were in the duodenal bulb. The mean lesion size was 10.1 (4.1) mm. With regard to macroscopic type, 12% of the lesions were classified as protruding (Paris classification type 0-I) and 88% as flat (type 0-IIa or 0-IIc). Preoperative biopsy had been performed in 54% of the patients.
Figure 2.
Flowchart of the participant selection. SNADET, Superficial nonampullary duodenal epithelial tumor; UEMR, underwater endoscopic mucosal resection.
Table 1.
Clinical characteristics of study participants (n = 367)
| Characteristic | Data |
|---|---|
| Age | |
| Mean (SD) | 63.5 (10.2) |
| Sex | |
| Male, n (%) | 241 (66) |
| Lesion size | |
| Mean (SD), mm | 10.1 (4.1) |
| Location | |
| Bulb, n (%) | 23 (6) |
| Superior duodenal angle, n (%) | 45 (12) |
| Descending part, n (%) | 224 (61) |
| Inferior duodenal angle, n (%) | 61 (17) |
| Transverse part, n (%) | 14 (4) |
| Location along circumference of duodenum | |
| Bulb | |
| Anterior aspect, n (%) | 9 (3) |
| Inferior aspect, n (%) | 4 (1) |
| Posterior aspect, n (%) | 6 (2) |
| Superior aspect, n (%) | 4 (1) |
| Others | |
| Anterior wall, n (%) | 45 (12) |
| Lateral wall, n (%) | 125 (34) |
| Posterior wall, n (%) | 104 (28) |
| Medial wall, n (%) | 70 (19) |
| Macroscopic type (Paris classification) | |
| Flat type (0-IIa, 0-IIc), n (%) | 324 (88) |
| Protruding type (0-I), n (%) | 43 (12) |
| Preoperative biopsy | |
| Performed, n (%) | 198 (54) |
| Treatment period | |
| Early (November 2017 to December 2020), n (%) | 183 (50) |
| Late (January 2021 to October 2024), n (%) | 184 (50) |
| Operator certification | |
| Board certified, n (%) | 304 (83) |
| Noncertified, n (%) | 63 (17) |
SD, Standard deviation.
Clinical outcomes of duodenal UEMR
The rates of en bloc resection and R0 were 88.8% and 65.1%, respectively. No intraoperative perforations occurred; intraoperative bleeding was observed in 1.9% of the patients. Delayed bleeding also occurred in 1.9% of patients. The mean procedural time was 6.4 minutes. Treatment outcomes stratified by operator certification status (board certified vs noncertified) are summarized in Table 2.
Table 2.
Treatment outcomes by certification status
| Outcome | Board certified (n = 304) | Noncertified (n = 63) |
|---|---|---|
| Resection fashion | ||
| En bloc resection, n (%) | 267 (88) | 59 (94) |
| Piecemeal resection, n (%) | 37 (12) | 4 (6) |
| Resection margin | ||
| R0, n (%) | 192 (63) | 47 (75) |
| Non-R0, n (%) | 112 (37) | 16 (25) |
R0, En bloc resection with histologically tumor-free horizontal and vertical margins.
Histologic outcomes
Histologically, 96% of the resected lesions were adenomas and 4% were intramucosal adenocarcinomas. Detailed information on resection margin status is presented in Table 3.
Table 3.
Histologic characteristics of resected lesions (n = 367)
| Histologic type | |
| Low-grade adenoma, n (%) | 285 (78) |
| High-grade adenoma, n (%) | 67 (18) |
| Intramucosal adenocarcinoma, n (%) | 15 (4) |
| Resection margin status | |
| R0 (HM−, VM−), n (%) | 239 (65) |
| R1 (HM+ and/or VM+), n (%) | 13 (4) |
| RX (indeterminate), n (%) | 115 (31) |
R0 (HM−, VM−) indicates that both the horizontal margin and vertical margin are negative. R1 (HM+ and/or VM+) indicates that the horizontal margin and/or vertical margin is positive.
HM, Horizontal margin; R0, en bloc resection with histologically tumor-free horizontal and vertical margins; R1, resection with positive horizontal and/or vertical margins; RX, resection with either the horizontal or vertical margin could not be assessed; VM, vertical margin.
Local recurrence
Of the 367 patients, follow-up data were available for 329 (89.6%). Local recurrence was observed in 9 of these 329 patients (2.7%). Among the 41 piecemeal resection cases, all had available follow-up, and 6 showed recurrence (14.6%). In contrast, among the 288 en bloc resection cases with follow-up, only 3 recurrences were observed (1.0%). Even within the en bloc resection group, 85 cases had non-R0, yet recurrence occurred in only 3 of these (3.5%).
Kaplan-Meier analysis of recurrence-free survival showed significantly poorer outcomes in the piecemeal and RX/R1 group than in the en bloc resection groups (log-rank test, P < .001; Wilcoxon test, P < .001). No significant difference in recurrence-free survival was observed between en bloc R0 and en bloc RX/R1 (Fig. 3).
Figure 3.
Kaplan-Meier curves of recurrence-free survival according to the resection method and histologic margin status. Solid line, en bloc R0 (R0: en bloc resection with histologically tumor-free horizontal and vertical margins); dashed line, en bloc R1/RX (R1, resection with positive horizontal and/or vertical margins; RX, resection with either the horizontal or vertical margin could not be assessed); and dotted line, piecemeal. Numbers at risk at 0, 1, 2, 3, 4, and 5 years are shown beneath the x-axis. Recurrence-free survival in the piecemeal group was significantly poorer than in both en bloc groups (log-rank test, P < .001; Wilcoxon test, P < .001), whereas no significant difference was observed between en bloc R0 and en bloc RX/R1. Censoring tick marks are omitted for clarity.
Predictors of piecemeal resection
We performed logistic regression analysis to identify the independent predictors of piecemeal resection in UEMR. In the univariate analysis, each 1-mm increase in lesion size and flat macroscopic type were significantly associated with an increased likelihood of piecemeal resection. Subsequent multivariate analysis identified each 1-mm increase in lesion size (odds ratio [OR], 1.20; 95% confidence interval [CI], 1.11-1.31) and flat macroscopic type (OR, 12.49; 95% CI, 2.41-238.05) as independent predictors of piecemeal resection (Table 4).
Table 4.
Predictors of piecemeal resection
| Variables | Univariate analysis |
Multivariate analysis |
||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P value | OR | 95% CI | P value | |
| Location | ||||||
| Nonbulb | 2.89 | 0.58-52.50 | .230 | — | — | — |
| Bulb | 1 | Reference | — | — | — | — |
| Lesion size | 1.16 | 1.08-1.26 | <.001 | 1.20 | 1.11-1.31 | <.001 |
| Macroscopic type (Paris classification) | ||||||
| Flat type (0-IIa, 0-IIc) | 5.92 | 1.23-106.24 | .022 | 12.49 | 2.41-238.05 | <.001 |
| Protruding type (0-I) | 1 | Reference | — | 1 | Reference | — |
| Biopsy | ||||||
| Present | 0.88 | 0.46-1.69 | .710 | — | — | — |
| Absent | 1 | Reference | — | — | — | — |
| Treatment period | ||||||
| Early (November 2017 to December 2020) | 1.19 | 0.62-2.29 | .606 | — | — | — |
| Late (January 2021 to October 2024) | 1 | Reference | — | — | — | — |
| Operator certification | ||||||
| Board certified | 2.04 | 0.78-7.02 | .156 | 1.58 | 0.58-5.54 | .398 |
| Noncertified | 1 | Reference | — | 1 | Reference | — |
“Reference” indicates the reference category for comparison in logistic regression.
—, Not included in the multivariate analysis on the basis of selection criteria; CI, confidence interval; OR, odds ratio.
Through ROC curve analysis, we determined that 10 mm was the optimal cutoff lesion size for predicting piecemeal resection, with an area under the curve of 0.692, a sensitivity of 80.5%, and a specificity of 72.0% (Fig. 4).
Figure 4.
Receiver operating characteristic curve of lesion size for predicting the risk of piecemeal resection. The area under the curve (AUC) was 0.692 and is marked with a star symbol (★) on the graph.
We also conducted logistic regression analyses to identify factors associated with non-R0. In the univariate analysis, each 1-mm increase in lesion size showed significant associations with non-R0. The subsequent multivariate analysis revealed that each 1-mm increase in lesion size (OR, 1.13; 95% CI, 1.07-1.19) and flat macroscopic type (OR, 2.19; 95% CI, 1.05-4.86) were independent predictors of non-R0 (Table 5).
Table 5.
Predictors of non-R0 (R1 or RX)
| Variables | Univariate analysis |
Multivariate analysis |
||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P value | OR | 95% CI | P value | |
| Location | ||||||
| Nonbulb | 1.22 | 0.49-2.85 | .661 | — | — | — |
| Bulb | 1 | Reference | — | — | — | — |
| Lesion size | 1.11 | 1.05-1.17 | <.001 | 1.13 | 1.07-1.19 | <.001 |
| Macroscopic type (Paris classification) | ||||||
| Flat type (0-IIa, 0-IIc) | 1.44 | 0.73-3.02 | .300 | 2.19 | 1.05-4.86 | .036 |
| Protruding type (0-I) | 1 | Reference | — | 1 | Reference | — |
| Biopsy | ||||||
| Present | 1.27 | 0.83-1.96 | .277 | 1.38 | 0.88-2.17 | .158 |
| Absent | 1 | Reference | — | — | Reference | — |
| Treatment period | ||||||
| Early (November 2017 to December 2020) | 1.15 | 0.75-1.76 | .536 | — | — | — |
| Late (January 2021 to October 2024) | 1 | Reference | — | — | — | — |
| Operator certification | ||||||
| Board certified | 1.71 | 0.95-3.25 | .077 | 1.43 | 0.77-2.76 | .260 |
| Noncertified | 1 | Reference | — | 1 | Reference | — |
“Reference” indicates the reference category for comparison in logistic regression.
—, Not included in the multivariate analysis on the basis of selection criteria; CI, confidence interval; OR, odds ratio; R0, en bloc resection with histologically tumor-free horizontal and vertical margins; R1, resection with positive horizontal and/or vertical margins; RX, resection with either the horizontal or vertical margin could not be assessed.
Discussion
We found that increasing lesion size and a flat macroscopic type are independent predictors of piecemeal resection in UEMR for SNADETs and that piecemeal resection was a predictor of local recurrence. ROC curve analysis revealed that 10 mm was the optimal cutoff lesion size for predicting piecemeal resection.
Several studies of the safety and technical feasibility of UEMR have been published.12,16,17,29, 30, 31 However, a major limitation of UEMR is that the rate of en bloc resection is lower than that with ESD,13 which may increase the risk of local recurrence.6,8,13,22, 23, 24, 25 Although most recurrent lesions can be treated with additional endoscopic treatment, some may necessitate salvage surgery.23,30 Repeated interventions, whether endoscopic or surgical, may elevate the risk of procedure-related adverse events and increase health care costs; thus, the need for additional treatment should be minimized. Although achieving histologic R0 is an important goal, our findings suggest that en bloc resection, even with non-R0, is associated with a low recurrence rate (1.0% for R0 and 3.5% for non-R0). In contrast, piecemeal resection demonstrated a markedly higher recurrence rate (14.6%), supporting the notion that fragmentation of the specimen may lead to residual disease irrespective of margin status. Kaplan-Meier analysis of recurrence-free survival further confirmed that the piecemeal group had significantly poorer outcomes than both en bloc resection groups, whereas no significant difference was observed between en bloc with R0 and en bloc with non-R0. These results underscore the clinical importance of avoiding piecemeal resection whenever feasible.
To date, the predictors of piecemeal resection in UEMR have not been evaluated systematically with robust statistical analysis. Our findings bridge this gap by providing quantitative evidence on the lesion characteristics that influence resection outcomes. The increased risk of piecemeal resection associated with the flat macroscopic type is consistent with findings from colorectal UEMR.32 One possible reason why achieving en bloc resection is challenging is that flat lesions lack a raised edge, making the snare more likely to slip off.32 Moreover, protruding lesions tend to be more buoyant than flat lesions when submerged, and their buoyancy enhances visibility and facilitates snare capture.
In a representative case of piecemeal resection during UEMR, the SNADET was a 16-mm flat (0-IIa) lesion on the lateral wall distal to the major papilla in the descending part (Fig. 5A). During UEMR for this lesion, the anal side of the lesion was poorly visualized under water (Fig. 5B). Although en bloc resection was attempted, snaring on the anal side was ineffective because of the lesion’s flat macroscopic type; as a result, piecemeal resection was necessary (Fig. 5C and D). An additional UEMR was subsequently performed for the residual lesion, and complete removal was confirmed visually (Fig. 5E). The mucosal defect was subsequently closed completely with endoscopic clips (Fig. 5F).
Figure 5.
A representative case of piecemeal resection during underwater endoscopic mucosal resection (UEMR). A, A 16-mm flat-type lesion located on the lateral wall distal to the major papilla in the descending duodenum. B, Underwater conditions resulted in poor visualization of the anal side of the lesion. C, Although en bloc resection was attempted, a residual lesion remained on the anal side (arrowheads). D, An additional UEMR was subsequently performed for the residual lesion. E, Complete resection was visually confirmed. F, The mucosal defect was completely closed with endoscopic clips.
Our analysis suggests that as the lesion size increases, particularly for flat lesions, UEMR alone is not always the optimal strategy. In recent years, PI-UEMR, which is a combination of UEMR with submucosal injection into only the distal edge of the lesion to improve visualization, has been reported to enhance the en bloc resection rate for SNADETs.33 Rates of en bloc resection are significantly higher with PI-UEMR (96%) than with standard UEMR (83%; P < .05), and the safety profiles of the 2 procedures are comparable.34 However, the evidence supporting PI-UEMR remains insufficient. Furthermore, ESD remains a viable option for select large or complex lesions, particularly when en bloc resection is crucial for minimizing local recurrence.13 Duodenal ESD, a technically demanding procedure, carries a higher risk of adverse events,14,35 which necessitates careful patient selection and meticulous procedural planning. Therefore, for moderately large or flat lesions, treatment strategies should be considered carefully to optimize resection outcomes.
This study has several limitations. First, its retrospective design may have introduced selection bias, because treatment decisions were made at the discretion of the endoscopist. Some large flat lesions in the duodenal bulb may have been excluded from the study because the endoscopist judged that the structural characteristics of the bulb would make en bloc snare capture more challenging under water. The resulting omission of technically demanding cases may have led us to underestimate the true risk for piecemeal resection in the duodenal bulb. Second, we did not assess how different types of snares may have affected overall procedural efficacy. Third, because this investigation was a single-center study, our findings may not be fully generalizable to institutions with varying levels of expertise and procedural experience.
In conclusion, our findings indicated that piecemeal resection was a risk factor for local recurrence in UEMR for SNADETs and that each 1-mm increase in lesion size and a flat macroscopic type were independent predictors of piecemeal resection. When performing UEMR for SNADETs, endoscopists should consider these factors to minimize the risk of local recurrence.
Patient consent
All patients provided written informed consent for the treatment before the procedure. This study adhered to the 2013 revision of the Declaration of Helsinki, and the study protocol was approved by our Institutional Review Board (approval no. 20190139). The requirement for written informed consent was waived by the Institutional Review Board because of the retrospective nature of the study.
Disclosure
The following authors disclosed financial relationships: T. Akimoto, T. Masunaga, Y. Takatori, N. Matsuura, A. Nakayama, K. Takabayashi, M. Kato, and N. Yahagi serve as consultants for Olympus Corporation. T. Akimoto, Y. Takatori, K. Miyazaki, N. Matsuura, M. Kato, and N. Yahagi have received honoraria as paid speakers from Olympus Corporation. T. Akimoto and M. Kato have received research funding from Olympus Corporation. M. Kato has received research funding from Boston Scientific. M. Kato and N. Yahagi have received honoraria as paid speakers from AMCO Inc. All other authors disclosed no financial relationships.
References
- 1.Jepsen J.M., Persson M., Jakobsen N.O., et al. Prospective study of prevalence and endoscopic and histopathologic characteristics of duodenal polyps in patients submitted to upper endoscopy. Scand J Gastroenterol. 1994;29:483–487. doi: 10.3109/00365529409092458. [DOI] [PubMed] [Google Scholar]
- 2.Arai T., Murata T., Sawabe M., et al. Primary adenocarcinoma of the duodenum in the elderly: clinicopathological and immunohistochemical study of 17 cases. Pathol Int. 1999;49:23–29. doi: 10.1046/j.1440-1827.1999.00820.x. [DOI] [PubMed] [Google Scholar]
- 3.Yoshida M., Yabuuchi Y., Kakushima N., et al. The incidence of non-ampullary duodenal cancer in Japan: the first analysis of a national cancer registry. J Gastroenterol Hepatol. 2021;36:1216–1221. doi: 10.1111/jgh.15285. [DOI] [PubMed] [Google Scholar]
- 4.Nakayama A., Kato M., Takatori Y., et al. How I do it: endoscopic diagnosis for superficial non-ampullary duodenal epithelial tumors. Dig Endosc. 2020;32:417–424. doi: 10.1111/den.13538. [DOI] [PubMed] [Google Scholar]
- 5.Nakagawa K., Sho M., Fujishiro M., et al. Clinical practice guidelines for duodenal cancer 2021. J Gastroenterol. 2022;57:927–941. doi: 10.1007/s00535-022-01919-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Apel D., Jakobs R., Spiethoff A., et al. Follow-up after endoscopic snare resection of duodenal adenomas. Endoscopy. 2005;37:444–448. doi: 10.1055/s-2005-861287. [DOI] [PubMed] [Google Scholar]
- 7.Lépilliez V., Chemaly M., Ponchon T., et al. Endoscopic resection of sporadic duodenal adenomas: an efficient technique with a substantial risk of delayed bleeding. Endoscopy. 2008;40:806–810. doi: 10.1055/s-2008-1077619. [DOI] [PubMed] [Google Scholar]
- 8.Alexander S., Bourke M.J., Williams S.J., et al. EMR of large, sessile, sporadic nonampullary duodenal adenomas: technical aspects and long-term outcome (with videos) Gastrointest Endosc. 2009;69:66–73. doi: 10.1016/j.gie.2008.04.061. [DOI] [PubMed] [Google Scholar]
- 9.Yamamoto Y., Yoshizawa N., Tomida H., et al. Therapeutic outcomes of endoscopic resection for superficial non-ampullary duodenal tumor. Dig Endosc. 2014;26 Suppl 2:50–56. doi: 10.1111/den.12273. [DOI] [PubMed] [Google Scholar]
- 10.Yahagi N., Kato M., Ochiai Y., et al. Outcomes of endoscopic resection for superficial duodenal epithelial neoplasia. Gastrointest Endosc. 2018;88:676–682. doi: 10.1016/j.gie.2018.05.002. [DOI] [PubMed] [Google Scholar]
- 11.Kinoshita S., Nishizawa T., Ochiai Y., et al. Accuracy of biopsy for the preoperative diagnosis of superficial nonampullary duodenal adenocarcinoma. Gastrointest Endosc. 2017;86:329–332. doi: 10.1016/j.gie.2016.12.007. [DOI] [PubMed] [Google Scholar]
- 12.Kiguchi Y., Kato M., Nakayama A., et al. Feasibility study comparing underwater endoscopic mucosal resection and conventional endoscopic mucosal resection for superficial non-ampullary duodenal epithelial tumor <20 mm. Dig Endosc. 2020;32:753–760. doi: 10.1111/den.13524. [DOI] [PubMed] [Google Scholar]
- 13.Kato M., Takeuchi Y., Hoteya S., et al. Outcomes of endoscopic resection for superficial duodenal tumors: 10 years' experience in 18 Japanese high volume centers. Endoscopy. 2022;54:663–670. doi: 10.1055/a-1640-3236. [DOI] [PubMed] [Google Scholar]
- 14.Pérez-Cuadrado-Robles E., Quénéhervé L., Margos W., et al. ESD versus EMR in non-ampullary superficial duodenal tumors: a systematic review and meta-analysis. Endosc Int Open. 2018;6:E998–E1007. doi: 10.1055/a-0579-9050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Binmoeller K.F., Shah J.N., Bhat Y.M., et al. “Underwater” EMR of sporadic laterally spreading nonampullary duodenal adenomas (with video) Gastrointest Endosc. 2013;78:496–502.e1. doi: 10.1016/j.gie.2013.03.1330. [DOI] [PubMed] [Google Scholar]
- 16.Okimoto K., Maruoka D., Matsumura T., et al. Utility of underwater EMR for nonpolypoid superficial nonampullary duodenal epithelial tumors ≤20 mm. Gastrointest Endosc. 2022;95:140–148. doi: 10.1016/j.gie.2021.07.011. [DOI] [PubMed] [Google Scholar]
- 17.Toya Y., Endo M., Yamazato M., et al. Resectability of underwater endoscopic mucosal resection for duodenal tumor: a single-center, retrospective pilot study. J Gastroenterol Hepatol. 2021;36:3191–3195. doi: 10.1111/jgh.15638. [DOI] [PubMed] [Google Scholar]
- 18.Lv X.-H., Luo R., Lu Q., et al. Underwater versus conventional endoscopic mucosal resection for superficial non-ampullary duodenal epithelial tumors ≤20mm: a systematic review and meta-analysis. Dig Liver Dis. 2023;55:714–720. doi: 10.1016/j.dld.2022.09.001. [DOI] [PubMed] [Google Scholar]
- 19.Lee J.G., Lee S.P., Jang H.J., et al. Underwater versus conventional endoscopic mucosal resection for superficial non-ampullary duodenal epithelial tumors: a systematic review and meta-analysis. Dig Dis Sci. 2023;68:1482–1491. doi: 10.1007/s10620-022-07715-1. [DOI] [PubMed] [Google Scholar]
- 20.Shichijo S., Uedo N., Takeuchi Y., et al. Underwater endoscopic mucosal resection of residual duodenal tumor. Endoscopy. 2019;51:E329–E330. doi: 10.1055/a-0919-4357. [DOI] [PubMed] [Google Scholar]
- 21.Pimentel-Nunes P., Libânio D., Bastiaansen B.A.J., et al. Endoscopic submucosal dissection for superficial gastrointestinal lesions: European Society of Gastrointestinal Endoscopy (ESGE) guideline – update 2022. Endoscopy. 2022;54:591–622. doi: 10.1055/a-1811-7025. [DOI] [PubMed] [Google Scholar]
- 22.Abbass R., Rigaux J., Al-Kawas F.H. Nonampullary duodenal polyps: characteristics and endoscopic management. Gastrointest Endosc. 2010;71:754–759. doi: 10.1016/j.gie.2009.11.043. [DOI] [PubMed] [Google Scholar]
- 23.Hoibian S., Ratone J.-P., Gonzalez J.-M., et al. Endoscopic mucosal resection of sporadic duodenal nonampullary adenoma: outcomes of 130 patients with a long-term follow up in two tertiary French centers. Ann Gastroenterol. 2021;34:169–176. doi: 10.20524/aog.2021.0581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tomizawa Y., Ginsberg G.G. Clinical outcome of EMR of sporadic, nonampullary, duodenal adenomas: a 10-year retrospective. Gastrointest Endosc. 2018;87:1270–1278. doi: 10.1016/j.gie.2017.12.026. [DOI] [PubMed] [Google Scholar]
- 25.Valerii G., Tringali A., Landi R., et al. Endoscopic mucosal resection of non-ampullary sporadic duodenal adenomas: a retrospective analysis with long-term follow-up. Scand J Gastroenterol. 2018;53:490–494. doi: 10.1080/00365521.2018.1438508. [DOI] [PubMed] [Google Scholar]
- 26.Miyazaki K., Nakayama A., Sasaki M., et al. Resectability of small duodenal tumors: a randomized controlled trial comparing underwater endoscopic mucosal resection and cold snare polypectomy. Am J Gastroenterol. 2024;119:856–863. doi: 10.14309/ajg.0000000000002634. [DOI] [PubMed] [Google Scholar]
- 27.Fujimoto K., Fujishiro M., Kato M., et al. Japan Gastroenterological Endoscopy Society Guidelines for gastroenterological endoscopy in patients undergoing antithrombotic treatment. Dig Endosc. 2014;26:1–14. doi: 10.1111/den.12183. [DOI] [PubMed] [Google Scholar]
- 28.Kato M., Uedo N., Hokimoto S., et al. Guidelines for gastroenterological endoscopy in patients undergoing antithrombotic treatment: 2017 appendix on anticoagulants including direct oral anticoagulants. Dig Endosc. 2018;30:433–440. doi: 10.1111/den.13184. [DOI] [PubMed] [Google Scholar]
- 29.Furukawa M., Mitoro A., Ozutumi T., et al. Efficacy of underwater endoscopic mucosal resection for superficial non-ampullary duodenal epithelial tumor. Clin Endosc. 2021;54:371–378. doi: 10.5946/ce.2020.147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yamasaki Y., Uedo N., Akamatsu T., et al. Nonrecurrence rate of underwater EMR for ≤20-mm nonampullary duodenal adenomas: a multicenter prospective study (D-UEMR study) Clin Gastroenterol Hepatol. 2022;20:1010–1018.e3. doi: 10.1016/j.cgh.2021.06.043. [DOI] [PubMed] [Google Scholar]
- 31.Yamasaki Y., Uedo N., Takeuchi Y., et al. Underwater endoscopic mucosal resection for superficial nonampullary duodenal adenomas. Endoscopy. 2018;50:154–158. doi: 10.1055/s-0043-119214. [DOI] [PubMed] [Google Scholar]
- 32.Chen C.-A., Ho C.-H., Hsieh P.-H. Evaluation of factors associated with en bloc colonic underwater endoscopic mucosal resection. Adv Dig Med. 2021;8:155–162. [Google Scholar]
- 33.Takatori Y., Kato M., Masunaga T., et al. Feasibility study of partial submucosal injection technique combining underwater EMR for superficial duodenal epithelial tumors. Dig Dis Sci. 2022;67:971–977. doi: 10.1007/s10620-021-06925-3. [DOI] [PubMed] [Google Scholar]
- 34.Takatori Y., Kato M., Masunaga T., et al. Efficacy of partial injection underwater endoscopic mucosal resection for superficial duodenal epithelial tumor: propensity score-matched study (with video) Dig Endosc. 2022;34:535–542. doi: 10.1111/den.14103. [DOI] [PubMed] [Google Scholar]
- 35.Fukuhara S., Kato M., Iwasaki E., et al. External drainage of bile and pancreatic juice after endoscopic submucosal dissection for duodenal neoplasm: feasibility study (with video) Dig Endosc. 2021;33:977–984. doi: 10.1111/den.13907. [DOI] [PubMed] [Google Scholar]